Generalized Multi-protocol Label Switching based Virtualization for Cloud Computing

Abstract

Generalized Multi Protocol Label Switching (GMPLS) or Multi-Protocol Lambda Switching is a new technology which produce amplification in Multi Protocol Label Switching (MPLS) to provide support network switching, space switching, and packet switching for time and wavelength. Thus, we can say that GMPLS is extension of MPLS which provides resilience and restoration by automatic switching. In this paper, we provide a Docker based simu-1ation environment which emulates a complex network and create insight functioning of GMPLS. The Docker based simulation environment is light-weight and scalable. It emulates large and complex network with minimal specifications. We have tested this Docker based GMPLS simulation testbed that depends on scaled network resources. Our method has shown tremendous improvement over the other virtualization methods. In this paper, we do a comparative study based on UML, Virtual Box, and Docker and found that Docker consumes very less resources when compared to UML and Virtual Box. In a system with 1 TB Hard disk space and 16 GB RAM, we can design and study a topology with 30–40 nodes using Virtual Box and a topology with 10–15 nodes using UML. Whereas, with Docker a topology with 300-nodes can be designed and tested. Thus scalability is limited by system specifications.